I would say you made a good choice. At least one other user (Wad Cheber) agreed with your assessment, as do I.
I don't believe there's any specific guidance laid out anywhere, but the edit rejection reasons give you some clues:

- Spam or vandalism is pretty straightforward: edits that introduce promotional content, or deface the post for no reasons, should be rejected
- No improvement whatsoever is more nuanced. Edits should aim to improve the post in some way; if it does not, then this reason applies. Personally I rarely use this one, although I do sometimes use it on edits "correcting" regional spelling
- Irrelevant tags is, for the most part, quite straightforward. Tags that are truly tangential (like adding star-wars to a harry-potter question) should not be added, and those edits should be rejected.In more difficult cases, this often becomes a way for users to tacitly vote against certain tag policies (character tags being an easy example)
- Clearly conflicts with the author's intent exists to remind us that we do enforce some notion of the sovereignty of the original poster. This one might be chosen in the case where an edit is too substantial for comfort
- Attempt to reply is quite straightforward, and I probably use it the most; I don't know if I'm just weird like that or what. Frequently users without the reputation to comment (or anonymous users) will use the "edit" feature to leave a comment, or even answer a question. This should be rejected swiftly
- Causes harm is the most nebulous, obviously. It largely comes down to when you know the edit is wrong, but it doesn't quite fit under any of the other categories
As you can perhaps tell, most of these involve judgement calls to some degree. As you gain more experience reviewing edits, you'll develop an intuition about what does and does not make a good edit.
If I had to give you some specific advice, it would be the following:
- Take a look at editing and suggested-edit. The questions in those tags will help give you a sense of what kind of specific things to watch out for
- Don't robo-review, by which I mean don't blindly accept or reject edits. There's no rush to review an edit, so take your time and look at all relevant information. Personally I like to pull up the original post page in a separate tab, so I can see the comments under the post I'm reviewing
- Don't be afraid to use the "skip" button. There's no penalty for skipping a review, and it's much better to do that than to commit to a review you're not sure about. A rule of thumb: If someone took to meta to protest your decision, could you justify it? If not, skip.